Leukotriene D₄ affects localisation of vinculin in intestinal epithelial cells via distinct tyrosine kinase and protein kinase C controlled events

Formation of epithelial barriers is highly dependent on physical interactions between the participating cells. More precisely, epithelial cells work together to develop a series of highly specific and morphologically well-defined structures, such as tight junctions, desmosomes and cell-cell adhesion junctions (Weiss et al., 1998). The adhesion junction is a specialised region of the plasma membrane, where cadherin molecules function as adhesion receptors that are linked to actin filaments in close contact with the plasma membrane (Itoh et al., 1997).

The cadherins constitute a family of transmembrane proteins that interact with each other at the extracellular surface through a Ca²⁺-dependent adhesion mechanism (Chitaev and Troyanovsky, 1998). The intracellular cadherin region interacts directly with one of two cytoplasmic proteins, either a β-catenin or a γ-catenin molecule, which in turn associates with an α-catenin molecule. The latter is responsible for coupling the cadherin-catenin complex directly or indirectly to the actin cytoskeleton (Chitaev and Troyanovsky, 1998; Calautti et al., 1998). The association of α-catenin with β]γ-catenins is vital for epithelial cell polarity and for the integrity of epithelial tissues (Huber et al., 1997).

In addition to β-or γ-catenin, α-catenin binds, via its COOH-terminal region, directly to vinculin (Weiss et al., 1998). Vinculin is a cytoskeletal protein that is closely associated with both cell-matrix interactions and cell-cell junctions. This protein has multi-ligand properties and has been found to interact with a number of microfilament-associated proteins, such as talin, α-actinin, paxillin and α-catenin, which reportedly bind to either the head or tail domains of vinculin (Jockusch et al., 1995). Studies have shown that an intramolecular head-to-tail interaction can lock the vinculin molecule into a folded configuration, in which its binding properties are impaired (Rüdiger et al., 1998; Bubeck et al., 1997; Johnson and Craig, 1995). Moreover, it is believed that vinculin can serve as a dynamic link between the cytoskeleton and cell surface adhesion molecules, such as integrins and E-cadherins (Tozeren et al., 1998). Vinculin has also been implicated in the control of adhesion and motility of fibroblasts (Westmeyer et al., 1990; Rodriguez Fernandez et al., 1993), PC12 cells (Varnum-Finney and Reichardt, 1994) and chick neuronal cells (Sydor et al., 1996). In support of this, the loss of vinculin in F9 cells has been found to result in rounded morphology and decreased adhesion (Grover et al., 1987). All of the cited findings suggest that vinculin is essential for normal cell attachment and spreading (Goldmann et al., 1996).

During an inflammatory reaction, the intestinal epithelial cells respond to inflammatory mediators such as leukotrienes (Samuelsson et al., 1987; Serhan et al., 1996). This is best illustrated by observations indicating that the intestinal barrier is disrupted in a number of inflammatory conditions, resulting in increased permeability of the mucosal barrier (Yamada et al., 1993).

Leukotrienes are arachidonic acid derivates that serve as key signalling substances in most inflammatory processes (Samuelsson et al., 1987; Serhan et al., 1996). In general, it is assumed that leukotriene-induced effects are mediated via interaction with a specific plasma membrane receptor, CysLT₁ (Lynch et al., 1999). We have previously shown (Sjölander et al., 1990; Thodeti et al., 2000), as have others (Pace et al., 1993; Hoshino et al., 1998), that leukotrienes, for example LTD₄, elicit several intracellular signals, including protein tyrosine phosphorylations, calcium elevations and PKC activation. We have also demonstrated that LTD4 causes a dramatic rearrangement of the actin cytoskeleton in intestinal epithelial cells (Massoumi and Sjölander, 1998). The latter finding is in line with data showing that leukotriene synthesis is both necessary and sufficient to induce cortical actin polymerisation and rounding up of A431 cells (Peppelenbosch et al., 1993). In Swiss 3T3 fibroblasts, application of leukotrienes has been reported to invoke actin polymerisation, probably through activation of the small G-protein Rho (Peppelenbosch et al., 1995). Taken together, these results demonstrate the ability of leukotrienes to induce actin rearrangements, although the signalling mechanism responsible for this effect is still unclear.

It is known that an intact intestinal barrier depends on cell-cell adhesion, and that such cooperation between intestinal epithelial cells is impaired by inflammatory reactions. In light of that, we performed experiments to determine whether LTD₄ influences the E-cadherin-catenin-vinculin adhesion complex that is involved in interaction between intestinal epithelial cells. Inasmuch as vinculin is a substrate for both tyrosine kinase and PKC, we focused on the effects of the inflammatory mediator LTD4 on the phosphorylation status and properties of vinculin.

The antibodies used and their sources were as follows: Alexa 488 goat anti-mouse IgG and Alexa 594 donkey anti-goat IgG, Molecular Probes Inc. (Eugene, OR, USA); anti-actin and anti-α-catenin, Santa Cruz Biotechnology (Santa Cruz, CA, USA); anti-β-catenin and anti-γ-catenin, and anti-phosphotyrosine PY-20, Transduction (Lexington, KY, USA); anti-active β1 integrin clone 12G10, Serotec (Oxford, UK); anti-β1 integrin clone P4C10, Life Technologies (Gaithersburg, MD, USA); anti-vinculin and anti-phosphotyrosine antibody 4G10, Upstate Biotechnology (Lake Placid, NY, USA); Mouse IgG (normal), peroxidase-linked goat anti mouse and rabbit anti-goat IgG, DAKO A/S (Glostrup, Denmark). LTD₄ was obtained from Cayman Chemical Co. (Ann Arbour, MI, USA), GF109203X and MAPT/AM was from Calbiochem-Novabiochem Co. (San Diego, CA, USA), PP1 was from Alexis Co. (Läufelfingen, Switzerland), Protein A Sepharose was from Amersham Pharmacia Biotech AB (Uppsala, Sweden) and Protein G Plus Agarose was from Oncogene Research (Cambridge, MA, USA). All other chemicals were of analytical grade and were obtained from Sigma Chemical Co. (St Louis, MO, USA).

The intestinal epithelial cell line Intestine 407 (Henle and Deinhardt, 1957) was cultured as a monolayer in 75-cm² flasks in Eagle’s basal medium supplemented with 15% new-born calf serum, 55 μg/ml streptomycin, and 55 i.u./ml penicillin. Cell cultures were kept at 37°C in a humidified environment of 5% CO₂ and 95% air. The cells, which exhibit typical epithelial morphology and growth, were regularly tested to ensure the absence of mycoplasma contamination.

The cells were cultured for 5 days, washed, and allowed to rest for at least 30 minutes in a physiologically balanced Ca²⁺medium (136 mM NaCl, 4.7 mM KCl, 1.2 mM KH₂PO₄, 1.2 mM MgSO ₄, 50 mM NaHCO ₃, 20 mM Hepes, 1.0 mM CaCl ₂, 5.5 mM glucose, pH 7.4). Thereafter, the cells were or were not stimulated with 40 nM LTD ₄ for different periods of time (10, 60 or 900 seconds). In some experiments, the cells were pre-incubated for 15 minutes with either the PKC inhibitor GF109203X (2 μM) or the Src family tyrosine kinase inhibitor PP1 (10 μM) before exposure to LTD ₄4. The stimulations were terminated by adding ice-cold lysis buffer (20 mM Tris, pH 7.4, 150 mM NaCl, 5 mM EDTA, 2 mM orthovanadate, 60 μg/ml phenylmethylsulfonyl fluoride, 4 μg/ml leupeptin, and 1% NP-40). To ensure satisfactory lysis, the cells were kept at 4°C for 1 hour in lysis buffer. The remaining attached cell debris was then scraped loose into the lysis buffer and homogenised 10 times on ice with a Dounce glass tissue grinder. The cell suspension was centrifuged at 200 g for 10 minutes at 4°C to remove cells and cell debris. Finally, the lysates were cleared by centrifugation at 10,000 g for 5 minutes. Prior to anti-vinculin immunoprecipitation the obtained lysates were also precleared with 10 μg pre-immune mouse IgG and 20 μl of Protein G PLUS Agarose for 30 minutes at 4°C and centrifuged at 10,000 g for 2 minutes. The protein content was determined and compensated for in all supernatants. The resulting lysates were either directly loaded on gels or immediately used in immunoprecipitations. In the latter case, portions (1.0 mg/ml protein) were incubated with 10 μg of an antibody to either vinculin, tyrosine-phosphorylated proteins, α-, β-or γ-catenin for 2 hours at 4°C, and 20 μl of Protein G PLUS Agarose or 40 μl of Protein A Sepharose (50% slurry) beads solution was added to each sample. After 1 hour, the beads were washed four times with the lysis buffer supplemented with 1% Tween 20. Portions of these precipitations as well as the complete cell lysates were then boiled in SDS-sample buffer for 10 minutes and loaded onto 8% polyacrylamide gels. The separated proteins were transferred onto PVDF membranes, which were blocked in 3% BSA/PBS and incubated for 1 hour at 25°C in a 1:500 dilution of a primary antibody in PBS with 3% BSA. The membranes were probed with a 1:5000 dilution of a secondary antibody coupled to horseradish peroxidase for 1 hour at 25°C and thereafter developed with an ECL kit (Amersham).

The cells were seeded onto glass coverslips and grown for 5 days; they were serum-starved for the last 16 hours. Thereafter, the cells were stimulated with 40 nM LTD ₄ or 100 nM TPA with or without pre-incubation with 2 μM GF109203X or 10 μM PP1 for 15 minutes in a tissue culture incubator at 37°C. The stimulations were terminated by fixing for 10 minutes at room temperature in 3.7% paraformaldehyde/PBS solution, after which the cells were permeabilised in 0.5% Triton X-100/PBS solution for 5 minutes. The coverslips were subsequently washed twice in PBS and incubated at room temperature in 3% BSA/PBS solution for 15 minutes. The cells were then stained for 1 hour with specific antibodies against vinculin (5 μg/ml), α-catenin (10 μg/ml), β-catenin (5 μg/ml) or γ-catenin (]μg/ml). Thereafter, the coverslips were washed six times in PBS and incubated with a 1:200 dilution (in blocking buffer) of either Alexa 488 goat anti-mouse IgG or Alexa 594 donkey anti-goat IgG secondary antibodies. As negative controls, we used normal rabbit IgG or normal mouse IgG. The coverslips were finally washed six times in PBS and mounted in fluorescent mounting medium (DAKO A/S). Samples were examined and photographed in a Nikon Eclipse 800 microscope, using a 60× objective. Images were recorded with a scientific-grade, charge-coupled device (CCD) camera (Hamamatsu, Japan) and subsequently analysed with HazeBuster deconvolution software (VayTek, Inc., Fairfield, CT, USA). Confocal images were examined using a Bio-Rad Radiance 2000 confocal laser scanning system with a Nikon microscope (model TE300) equipped with a 60×]1.4 Plan-apochromat oil immersion objective.

cDNA coding for full-length human PKC-α or the regulatory domain of PKC, fused to EGFP cDNA (Zeidman et al., 1999) were obtained from Dr Christer Larsson. Transient transfections of the cells were achieved using 3.5 μl of Lipofectin (Gibco) and 1.8 μg of plasmid DNA/ml, and were performed in serum-free medium, essentially according to the protocol provided by the supplier.

Adhesion substrate was prepared by coating flat-bottomed 96-well microtitre plates (Nunc) overnight at 4°C with 50 μl of collagen IV, laminin or fibronectin (10 μg/ml), or for 5 minutes at room temperature with poly-L-lysine (10 μg/ml). Thereafter, the plates were washed with PBS, and the coated wells were blocked with 1% heat-inactivated BSA in PBS for 1 hour at 37°C and then washed three times with PBS. Confluent cells were serum-starved overnight and then detached by exposure to 0.05% trypsin/EDTA; trypsin activity was subsequently neutralised with Eagle’s basal medium supplemented with 15% newborn calf serum. The cells were collected by centrifugation and then resuspended in 2% BSA in serum-free Eagle’s basal medium. The cell suspension was incubated in a rotator (to avoid adherence) with or without inhibitors or antibodies (concentrations indicated in the legends to the figures) for 45 minutes at 37°C. Thereafter, cells (5×10 ⁵/well) were added to the coated wells and incubated for 1 hour at 37°C. The unattached cells were removed by gently washing three times with serum-free Eagle’s basal medium, and residual attached cells were stained using the MTS dye reduction assay (Promega, Madison, WI, USA). The cells were subsequently incubated for 3 hours at 37°C with a tetrazolium salt [3-(4,5-dimethyl thiazol-2-yl)-2.5-diphenyl tetrazolium bromide]. The metabolically active cells reduced the dye to purple farmazan, and absorbance at 490 nm was measured using a microplate reader (model BMG, Offenberg, Germany). Cells plated in wells coated with 3% BSA/PBS were used as controls.

Expression of integrins on the surface of untreated or LTD ₄-treated cells was determined by flow cytometry. Cells were harvested with 1 mM EDTA/EGTA in PBS, resuspended at a concentration of 4×105 cells in 150 μl of PSB solution (0.1% sodium azide and 1% BSA/PBS), and then incubated with 5 μg/ml anti-β1 integrin 12G10 or mouse IgG for 45 minutes on ice. Cells were washed with PSB, and incubated with fluorescein-conjugated secondary antibody Alexa 488 goat anti-mouse IgG (diluted 1:75) for 45 minutes on ice. Thereafter, the cells were washed again and fixed with 1% paraformaldehyde/PBS, and cell-bound fluorescence was quantified using a Coulter Epics FACScan flow cytometer.

Vinculin has been reported to be a target for tyrosine phosphorylation both in vivo and in vitro (Werth et al., 1983; Werth and Pastan, 1984), thus we performed experiments to determine whether LTD4 would induce tyrosine phosphorylation of vinculin in Int 407 cells. Immunoprecipitation using anti-vinculin or phosphotyrosine PY-20 antibodies followed by immunoblotting with anti-phosphotyrosine 4G10 or vinculin antibodies showed significant tyrosine phosphorylation of vinculin after 60 seconds of stimulation with LTD₄ (Fig. 1A,B). That effect was abolished by pretreatment with PP1 (10 μM), a specific inhibitor of Src-like family tyrosine kinases (Fig. 1A,B). In addition, blots analyzing anti-vinculin immunoprecipates were also reprobed with an anti-vinculin antibody to ensure equal loading of the different samples (Fig. 1A). For comparison we have also included an anti-phosphotyrosine blot of complete cell lysates of non-stimulated cells (Fig. 1A, Lys).

The activation of vinculin has been shown to increase its affinity for actin (Johnson and Craig, 1995). Therefore, we examined if the tyrosine phosphorylation of vinculin affected its binding to actin. Immunoprecipitation of vinculin and subsequent immunoblotting against actin revealed that treatment with LTD₄ led to increased vinculin-actin association (Fig. 2A,B). To ascertain whether this effect was due to the observed tyrosine phosphorylation of vinculin, we studied the influence of PP1 (10 μM) on LTD₄-induced vinculin-actin binding. The results show a clear correlation between impaired tyrosine phosphorylation of vinculin and annulment of vinculin-actin binding (Fig. 2A,B).

Exposing the cells to LTD₄ (40 nM) induced a prompt dissociation of vinculin from α-catenin, as revealed by immunoprecipitation of α-catenin followed by immunoblotting with an anti-vinculin antibody and as a control reprobing against α-catenin (Fig. 3A,B). This dissociation was detected 10 seconds after stimulation with LTD₄ and reached a plateau between 1 and 15 minutes of treatment. The observed dissociation was confirmed when we obtained the same results by immunoprecipitating vinculin and subsequently immunoblotting with an anti-α-catenin antibody, and as a control reprobing against vinculin (Fig. 3C,D). These findings suggest that activation of the LTD₄-Cys LT₁ receptor induces a prompt dissociation of vinculin from the catenin adhesion complex.

It has been shown that α-catenin links the cadherin-catenin complex to the cytoskeleton (Huber et al., 1997), thus we examined the effects of the inflammatory mediator LTD₄4 on this complex. In immmunoblot analysis, the same amount of β-or γ-catenin was recovered in α-catenin immunoprecipitates from unstimulated and LTD₄-treated cells (Fig. 4A), and identical results were obtained for α-catenin recovered in β-or γ-catenin immunoprecipitates (Fig. 4B). Uneven protein loadings were excluded by stripping and reprobing the blots with an antibody against the immunoprecipitated protein (Fig. 4A). The interaction of α-, β- and γ-catenin induced by exposure to LTD₄4 was also analysed by immunofluorescence microscopy. In double-staining experiments, α-catenin and β-catenin, as well as α-catenin and γ-catenin, exhibited very similar distribution in unstimulated and LTD₄-stimulated cells (data not shown).

To determine whether tyrosine phosphorylation of vinculin could also explain the LTD4-induced dissociation of vinculin from α-catenin, we preincubated cells for 15 minutes with 10 μM PP1. This treatment had no effect on the dissociation (Table 1). However, like LTD4, the phorbol ester TPA caused dissociation of vinculin from α-catenin (Table 1), and pretreatment with 2 μM GF109203X for 15 minutes significantly inhibited both the LTD4- and the TPA-induced dissociation (Table 1). These results suggest that PKC, but not a Src tyrosine kinase, is involved in separation of vinculin from α-catenin.

Vinculin distribution in control cells and cells stimulated with LTD₄ or TPA for 15 minutes in the presence or absence of GF109203X or PP1 was determined by fluorescence microscopy. As shown in Fig. 5, vinculin was homogeneously distributed in unstimulated cells, whereas stimulation with either LTD4 or TPA led to concentration of the protein in focal adhesions at the cell periphery. Futhermore, we also found an increased amount of paxillin in anti-vinculin immunoprecipitates from LTD4 stimulated cells (Fig. 6). Both the LTD4- and the TPA-induced accumulation of vinculin in focal adhesions was blocked by pretreatment with 2 μM GF109203X or 10 μM PP1 for 15 minutes (Fig. 5). Interestingly enough, the phorbol ester-induced accumulation of vinculin in focal adhesions was not completely blocked by PP1. This could possibly be due to the much stronger activation of PKC by the phorbol ester compared with that of a natural ligand such as LTD₄.

We have recently shown that LTD₄ activates PKC-α, as revealed by translocation of the enzyme to the plasma membrane (Thodeti et al., 2001). Therefore, we investigated the possibility that activation of PKC-α by LTD4 is involved in the dissociation of vinculin from α-catenin. We observed that the leukotriene caused an increased recovery of PKC-α in vinculin immunoprecipitates (Fig. 7), but we found no association between vinculin and any of the other isoforms of PKC (data not shown). We also noted that the PKC inhibitor GF109203X reduced the LTD4-induced recovery of PKC-α in anti-vinculin immunoprecipitates. Finally, we examined the influence of LTD4 on the distribution of vinculin and PKC-α in cells transfected with EGFP-PKC-α or with the regulatory domain of PKC (PKC-RD), demonstrated as an ideal dominant-negative inhibitor of PKC (Jaken, 1996). Vinculin and PKC-α were distributed fairly homogeneously throughout the cytoplasm of unstimulated cells (Fig. 8, top three panels) and were colocalised to focal adhesions in cells treated with LTD4 (Fig. 8, middle three panels), while the LTD4-induced accumulation of vinculin in focal adhesions was effectively blocked in cells transfected with EGFP-PKC-RD (Fig. 8, bottom three panels). Quantitative analysis of the staining and transfection reveals low values for the colocalisation coefficients in unstimulated cells for red: 0.10 (Texas Red, vinculin) and for green: 0.22 (EGFP-PKC-α). In the presence of LTD4 an increased degree of colocalisation was shown by coefficient values of 0.76 for red and 0.65 for green.

Cells treated with 40 nM LTD₄ for 15 or 30 minutes significantly increased their adhesion to collagen IV. In contrast, adhesion to laminin, fibronectin or poly-L-lysine, was not significantly affected by LTD₄ stimulation (data not shown). The adhesion of cells to collagen IV was prevented by pretreatment with an anti-β1 blocking antibody, P4C10, whereas treatment with normal mouse IgG (control) had no effect (Fig. 9). This LTD4-induced increase in cell adhesion was observed after stimulation for 15-30 minutes (Fig. 9). These results suggest that the LTD4-induced adhesion to collagen is mediated by β1-integrins. Fig. 9 also shows that the PKC inhibitor GF109203X, and the Ca²⁺ chelator MAPT/AM blocked the LTD₄-induced adhesion to collagen, suggesting that PKC is involved in the LTD₄-induced β1-dependent attachment of intestinal cells to collagen IV.

One possible explanation for the LTD₄-induced increase in cell adhesion is an elevation of active β1 integrins on the cell surface induced by intracellular signalling. To examine this, we investigated the effect of LTD4 on the cell surface expression of active β1 integrin by FACS analysis. Fig. 10 shows that when cells were treated with LTD₄ for 1 hour, we could detect an increased cell surface binding of anti-β1 integrin antibodies (clone 12G10), which only bind to the activated form of this integrin (Mould et al., 1995; Mould et al., 1998).

We found that the inflammatory mediator LTD₄ triggered a rapid tyrosine phosphorylation of vinculin in intestinal epithelial cells. It has been documented that vinculin can be phosphorylated on tyrosine, as well as on serine/threonine residues (Werth et al., 1983; Goldmann et al., 1996; Schwienbacher et al., 1996). So far very little has been reported about the functional consequences of vinculin tyrosine phosphorylation. However, in the present study we demonstrate that tyrosine phosphorylation of vinculin is part of the regulation of its interaction with actin in focal adhesions. Furthermore, our observation is interesting when considering that intestinal inflammation is associated with an impaired intestinal barrier, and vinculin is located in adhesion junctions, which mediate interactions between epithelial cells that are essential for an intact intestinal barrier.

In vitro studies show that vinculin can interact directly with α-catenin (Rüdiger, 1998; Hazan et al., 1997), and it also appears to participate in reorganisation of the cytoskeleton by direct interaction with filamentous actin (for reviews, see Goldmann 1996; Hüttelmaier et al., 1997; Johnson and Craig, 1995). In our study, LTD₄ caused rapid dissociation of vinculin from α-catenin without affecting the catenin complexes per se and the time kinetics of that effect were similar to that of the LTD₄-induced tyrosine phosphorylation of vinculin. In addition, LTD₄ increased the association between vinculin and actin, again with time kinetics similar to that of vinculin tyrosine phosphorylation. However, we found that PP1, an Src-like tyrosine kinase inhibitor, blocked LTD₄-induced tyrosine phosphorylation of vinculin and coupling of vinculin to actin, whereas it did not affect the dissociation of vinculin from α-catenin. These data suggest that an Src-like kinase is responsible for tyrosine phosphorylation of vinculin and that this modification of vinculin is required for coupling of vinculin to actin. The findings also indicate that some other mechanism(s) is involved in LTD₄-induced dissociation of vinculin from α-catenin.

Vinculin and α-catenin are both important elements in the formation of adhesion junctions, and therefore altered coordination of these proteins will affect the cell-cell interaction mediated by adhesion junctions (Aplin et al., 1999). Our results imply that PKC plays a role in dissociation of vinculin from α-catenin. This assumption is based on the following findings: TPA mimicked LTD₄ in inducing dissociation of vinculin from α-catenin; GF109203X (a protein kinase C inhibitor) blocked both the LTD₄- and the TPA-induced dissociation; and LTD₄ elicited a rapid association, direct or indirect, between vinculin and PKC-α, as revealed by immunoprecipitation experiments. We did not detect any other PKC isoforms in anti-vinculin immunoprecipitates. Consequently, our data indicate that a PKC-α-induced serine/threonine phosphorylation of vinculin regulates its dissociation from the catenin complexes. Regarding the activation of PKC, we have in a recent study demonstrated (Grönroos et al., 1996; Thodeti et al., 2000; Thodeti et al., 2001), that LTD₄ can activate phospholipase C and PKC. In support of those findings, Hyatt et al. (Hyatt et al., 1994) and Weekes et al., (Weekes et al., 1996) have reported that PKC-α interacted with vinculin in experiments in vitro. Moreover, in fibroblasts, phorbol esters cause codistribution of vinculin and PKC-α to focal contacts (Jaken et al., 1989). We extend these observations by demonstrating that the natural agonist LTD4, known to trigger intracellular signalling from a seven-transmembrane receptor, can induce an association between PKC-α and vinculin.

Our results demonstrating a coupling of vinculin to actin did not indicate where in the cells this interaction occurs, thus we performed immunofluorescence microscopy to determine the effect of LTD₄ on the distribution of vinculin. It was evident that the leukotriene induced accumulation of vinculin in focal adhesions, and the effect was blocked by transfection of PKC-RD (a dominant negative inhibitor of PKC; Jaken, 1996) and by preincubation with either the protein kinase C inhibitor GF109203X or the Src-like tyrosine kinase inhibitor PP1. However, Levenberg et al. (Levenberg et al., 1998) reported that fibronectin engagement of integrin receptors on CHO cells led to a reduction of vinculin in cell-cell adhesions but an increased amount of the protein in focal contacts. Furthermore, by expressing EGFP-PKC-α, we found that LTD4 caused vinculin to colocalise with PKC-α, and that effect was mimicked by the phorbol ester TPA.

It is clear that focal adhesions are dynamic structures that change in size, and most likely also in their composition, as the cell adhesion process progresses (Aplin et al., 1999), and that different integrins assemble different types of adhesion complexes (Katz et al., 2000). We found that the LTD4-induced increase in cell adhesion to collagen is mediated by β1 integrins and accompanied by translocations of PKC-α and vinculin to focal adhesions. We also studied the LTD4-induced inside-out signalling responsible for the regulation of β1 integrin activity. Our data suggest that a PKC, possibly PKC-α] is important in this signalling pathway, since both GF109203X and MAPT/AM prevented the LTD4-induced adhesion of Int 407 cells to collagen.

The formation of focal adhesions could compensate for the reduction in adhesion junctions and thereby maintain proper attachment of the intestinal cells to the extracellular matrix; a similar response has been observed in endothelial cells (Lewalle et al., 1997). More specifically, these investigators reported that interaction between breast adenocarcinoma cells (MCF-7) and normal endothelium induced a loss of interendothelial contacts and translocation of vinculin from cell borders to focal adhesion sites located on the basal plasma membrane. In addition to a role in cell attachment, focal adhesions have also been shown to mediate cell survival signals in many different types of cells (Aplin et al., 1999; Farrelly et al., 1999). Therefore, it is possible that LTD₄-induced formation of focal adhesions is also involved in the anti-apoptotic effect of LTD₄ on intestinal epithelial cells, that was recently reported by Öhd et al. (Öhd et al., 2000).

In conclusion, we found that the inflammatory mediator LTD4 induces PKC-dependent dissociation of vinculin from catenin complexes and simultaneous Src-like tyrosine kinase-mediated coupling of vinculin to actin. Both these signalling events are essential for LTD₄-induced accumulation of vinculin in focal adhesions, which is accompanied by an increased β1 integrin-dependent adhesion to collagen.

We are grateful to Dr Christer Larsson (Division of Molecular Medicine, Malmö University Hospital, Malmö, Sweden) for providing the full-length EGFP-PKC-α and EGFP-PKC-RD and Ms Patricia Ödman for linguistic revision of the manuscript. This work was supported by grants to A.S. from the Swedish Medical Research Council (project no. 10356), the Magn. Bergvalls Foundation, the Crafoord Foundation, Inga and John Hains’ Foundation, the Foundations at Malmö University Hospital, the Åke Wiberg Foundation, and the Östlunds Foundation and by grants to R.M. from the Royal Physiographic Society in Lund.